Show Don't Tell: What It Actually Means in 2026

"Show, don't tell" is the most-quoted writing advice in English. It is also the most-misunderstood. Working novelists hear it from beta readers, workshop peers, and developmental editors so often that the phrase has lost its meaning, and writers contort their drafts trying to follow a rule that was never the rule.

Short version, before the long one. Show don't tell is not a ban on telling. It is a rule about emphasis: the moments that matter most to the reader's experience of the story should be rendered as scene, not summary. Telling is fine, often necessary, for everything else. A novel that follows "show don't tell" literally is unreadable.

This post unpacks what the phrase actually means, where it came from, when telling is the correct move, and how to diagnose whether your draft is leaning too far in either direction.

The phrase, decoded

"Show don't tell" is shorthand for a more useful instruction: render the moments of greatest emotional or narrative weight as scenes that the reader experiences, not as summaries the narrator hands them.

The distinction is between scene and summary.

A scene is a discrete moment rendered in real time, with sensory detail, dialogue, beats, and the reader inhabiting the protagonist's experience. Time slows down. The reader is present.

A summary is the narrator stepping back and telling the reader what happened, often compressing weeks or months into a sentence. Time speeds up. The reader observes from outside.

Both are necessary. A 90,000-word novel that's all scene is exhausting and bloated. A 90,000-word novel that's all summary is a Wikipedia article. The craft is in choosing which is which.

The correct version of the rule is: scene where it matters, summary where it doesn't. "Show don't tell" tries to compress that into three words and ends up sounding like an absolute, which it never was.

Examples of show vs tell

The cleanest way to see the difference is to put them next to each other. Same beat, two ways.

Tell:

Marcus was angry with his father.

Show:

Marcus closed the door slowly. He had practiced this. The handle clicked into the strike plate the way his father's voice had clicked, decades earlier, when the answer was no and there would be no further discussion.

The first version is summary. Six words. The reader receives information.

The second version is scene. The reader is in Marcus's body, watches his hand on the door, hears the click, and inherits the connection between the present click and the historical voice. The reader feels the anger because the prose is doing the feeling. Forty words instead of six, but they're forty different words.

Now consider the inverse. Imagine the second version is your opening paragraph for an entire chapter about Marcus's day. He goes to work. He has lunch. He drives home. He sees his father at dinner. Each of those should also be rendered with the same forty-word density?

No. That book is 600 pages and unreadable.

What the second version actually does is announce: this is the beat that matters. The door, the click, the historical resonance. After the door, Marcus drives to work in two sentences. He has lunch in one. He drives home in a paragraph. The reader doesn't need scene-level rendering for those, because nothing in them is doing emotional or narrative work that the protagonist is going to act on later.

That's the rule. Render scene at the points where the reader needs to inhabit the experience to feel what the protagonist is feeling. Summarize the rest.

When telling is the right move

There are at least five places where telling beats showing.

1. Time compression

If a year passes between two scenes and nothing in that year matters to the story, you tell.

Three years passed. Marcus finished medical school. His father died in the second year, alone in a hospital his son did not visit.

That's twenty words. Rendering it as scene would take 30,000 words and dilute the next moment that actually matters. Compression is the reader's friend.

2. Logistical glue

Getting a character from point A to point B is rarely scene-worthy. If the trip itself doesn't carry weight, summarize it.

She took the train down on Friday morning.

Rather than:

The train pulled out of King's Cross at 7:42, three minutes late, and she watched the city slide past the window for the first hour, then opened her book...

Unless the train ride is doing work (revelation, confrontation, mood-setting that the chapter needs), summary is correct.

3. Establishing context

Some background information is necessary and not improvable by dramatization. World rules, historical context, family structure: these can sometimes be told.

The rule had been simple for two centuries. The crown went to the eldest male. Then, in the seventh year of the war, it didn't.

Trying to dramatize "the rule had been simple for two centuries" through scene work is often worse than telling the reader plainly.

4. Interior summary

When a character has reached a conclusion through scenes you've already rendered, it is okay to summarize the conclusion.

She knew, then, that her marriage was over.

You don't need to dramatize the realization if the reader has been watching the marriage erode for sixty pages. The reader is ahead of the character. Tell.

5. Voice

Sometimes the narrator's telling is the voice. The Catcher in the Rye is mostly Holden telling. The Goldfinch opens with seventeen pages of Theo telling. Gatsby is Nick telling. The line between scene and summary collapses when the voice itself is the show.

If a literary writer with a strong narrative voice is told to "show, don't tell" in a workshop, the right answer is to ignore the note. The voice is doing the showing.

When showing is the right move

Conversely, here are the places where you must show. These are non-negotiable in any genre.

1. The inciting incident

The event that pulls the protagonist out of their normal life cannot be summarized. The reader has to live it.

2. The protagonist's first significant choice

Choices reveal character. Summarizing a choice ("he decided to go") flattens the moment. Showing a choice ("his hand was already on the door before he'd finished the thought") lands the character.

3. Major emotional beats

First kiss, death of a loved one, public humiliation, betrayal by a parent, the moment the secret comes out. These are the moments the reader buys the book for.

4. The midpoint

The structural fulcrum of a three-act novel. If your midpoint is summarized, the second half of the book has nothing to push against.

5. The climax

The final confrontation. If your reader has read 250 pages to get to the climax and you summarize the climax in three paragraphs, you have broken the contract.

6. Voice-defining moments

Beats where the protagonist's specific way of being in the world is on display. If your protagonist has a particular relationship with silence, give us a silence-rendered scene early. Telling us "she was a quiet person" is not a substitute.

How to diagnose your own draft

The fastest way to see your show/tell ratio is a 30-minute exercise.

1. Open your manuscript.
2. Pick the chapter you suspect is weakest.
3. Take a highlighter and mark every paragraph as either S (scene) or T (tell).
4. Count.

If your weakest chapter is 90% T, you're handing the reader information instead of dramatizing it. The fix is to identify the two or three beats that should have been scenes and re-render them.

If your weakest chapter is 95% S, you're spending 4,000 words rendering a beat that didn't need 4,000 words. The fix is compression.

The strongest chapters in commercial fiction tend to run 70% scene and 30% tell, with the tells doing transitions, time compression, and interior summary. Literary fiction can run higher on either end. Voice-forward literary novels sometimes run 60% tell and still work because the voice is the show.

The "filtering" trap

A common middle-ground failure is filtering. Filtering happens when the writer thinks they're showing but is actually telling at one remove.

Filtered scene:

Marcus saw his father walk into the kitchen. He noticed how slowly his father moved. He felt a wave of pity rise in his chest.

That's three filtering verbs (saw, noticed, felt) inserting the protagonist between the reader and the experience. The reader is observing Marcus observing the father, not experiencing the father directly.

Unfiltered scene:

His father walked into the kitchen, slowly. The pity rose like a wave.

The reader is now in Marcus's body. The filter verbs are gone. The image lands.

If your prose is technically scene but reads flat, you're often filtering. Cut the saw, watched, noticed, felt, heard, smelled, realized, and the prose snaps closer to the reader.

The "show don't tell" feedback loop

If you're getting the note in workshop or from beta readers, the cause is almost always one of these:

1. You're summarizing a beat the reader needed to inhabit. Promotion: turn the summary into a scene.
2. You're filtering. Unfilter the prose.
3. You're telling the reader what to feel. ("She was devastated.") Trust the reader. Show the cause; let them feel the effect.
4. Your scene is doing tell-work. You wrote a scene but the dialogue is exposition, the action is functional, and the reader is still being told. Rewrite the scene around the emotional beat instead of the plot information.

The fix is rarely "more showing everywhere". It's usually "show the right beats, summarize the rest, and stop filtering".

Where AI fits

Most AI writing tooling in 2026 doesn't help with show vs tell because the diagnostic requires reading the whole chapter and identifying which beats deserved scene-level rendering. That's a structural read, not a sentence-level one. Tools like Grammarly and ProWritingAid can flag passive voice and "to be" constructions, which sometimes correlate with telling, but the correlation is loose. A great tell can have zero passive voice.

A small number of tools, including Inkett's Editor, run a chapter-level pass that flags beats where the prose appears to summarize a moment that, given its position in the structure, probably warranted scene-level rendering. It also flags filtering at the line level. The output reads like the margin notes a $5,000 freelance editor would write, anchored to specific paragraphs.

The point isn't replacement. Show vs tell is a craft judgment that depends on knowing your own book. The AI is useful as a second opinion when you're too close to the manuscript to see the shape clearly.

Closing

"Show don't tell" is good advice misused as a hammer. The real instruction is: render the moments that matter as scenes, summarize the rest, and stop filtering when you mean to be in the protagonist's body. Run the highlighter exercise on your weakest chapter and the diagnosis will name itself.

Inkett is the writing stack for working novelists. The Editor catches the show/tell imbalance and the filtering at the chapter level so you can revise the right paragraphs instead of guessing. The voice stays yours.